Customizing Backend AI Models: Optimizing Performance

Learn how to customize which AI models power different components of LiGo, allowing you to optimize generation speed and quality based on your specific needs

Last updated: March 18, 2025
5 min read
StandardPremium

LiGo gives you the flexibility to customize which AI models power different components of the extension, allowing you to optimize for speed, quality, or specific writing styles. This guide explains how to configure these settings for your needs.

What You'll Learn

  • How to access AI model selection settings
  • Customizing models for different content types
  • Optimizing generation speed vs. quality
  • Managing auto-generation settings
  • Recommended configurations for different use cases

Accessing Model Selection Settings

Model Selection

To customize your AI model settings:

  1. Click the LiGo extension icon in your browser
  2. Select "Settings" (gear icon) in the bottom of the sidebar
  3. Navigate to the "Generation Settings" section
  4. Here you'll find options for customizing the AI models for different components

Alternative Access: You can also find these settings in the web app by going to Settings and selecting the Generation tab.

Understanding Model Options

LiGo allows you to select from different AI models for each component:

Available Models

  • Claude: Balanced quality and speed with excellent nuance in professional contexts
  • ChatGPT: Fast generation with strong performance across various content types
  • Additional Models: LiGo periodically adds new model options as AI technology evolves

Each model has different strengths, and the optimal choice may depend on your specific content needs and preferences.

Customizing Models by Component

Comment Generation

You can select different models for:

  1. LiGo Style Comments:

    • Click the dropdown next to "Generate 'LiGo' style comments"
    • Select your preferred model (e.g., ChatGPT, Claude)
    • Toggle the switch ON to enable this component
  2. My Style Comments:

    • Click the dropdown next to "Generate 'My Style' comments"
    • Select your preferred model
    • Toggle the switch ON to enable this component

Pro Tip: You might prefer different models for different comment styles. For example, you might find Claude better captures your personal voice for custom styles, while ChatGPT performs well for LiGo's optimized style.

Post Generation

Similarly, you can customize models for post generation:

  1. LiGo Style Posts:

    • Click the dropdown next to "Generate 'LiGo' style posts"
    • Select your preferred model
    • Toggle the switch ON to enable this component
  2. My Style Posts:

    • Click the dropdown next to "Generate 'My Style' posts"
    • Select your preferred model
    • Toggle the switch ON to enable this component

Note: "My Style" refers to your selected default style pack. If you've renamed your style, the UI will still show "My Style" in the settings page.

Auto-Generate Settings

LiGo offers an auto-generation option for more streamlined workflows:

  • Auto-generate comments: Toggle ON to automatically generate comments when clicking the "Engage" button
  • When enabled, the extension will immediately start generating comments when you open it on a LinkedIn post
  • When disabled, you'll need to manually initiate comment generation

Pro Tip: Enable auto-generation for faster workflows on frequently visited LinkedIn profiles. Disable it when browsing to avoid unnecessary processing.

Performance Optimization Strategies

To optimize LiGo's performance for your specific needs:

For Maximum Speed

  • Disable any unused component toggles
  • Select ChatGPT for components where speed is prioritized over nuance
  • Limit enabled style packs to just those you use regularly

For Best Quality

  • Select Claude for components where nuance and professional tone are most important
  • Enable both LiGo and My Style options to get more variations to choose from
  • Keep auto-generation enabled for a smoother workflow

For Balanced Performance

  • Use ChatGPT for comment generation (usually faster)
  • Use Claude for post generation (usually more nuanced)
  • Disable any styles you use infrequently

Understanding Model Differences

Different models may produce slightly different content styles:

  • Claude often excels at capturing nuanced professional tones and adapting to your personal style examples
  • ChatGPT typically generates content quickly and performs well with structured content formats
  • Model performance may vary based on your specific industry, voice, and content type

Pro Tip: Experiment with different models for each component to find the combination that best matches your voice and workflow needs.

Saving Your Settings

Once you've customized your model selections:

  1. Review your choices for each component
  2. Click the "Save Settings" button at the bottom of the page
  3. Your preferences will be applied immediately for future content generation

Settings will persist across sessions and synchronize between devices when using the same LiGo account.

Model Availability Notes

A few important points about model availability:

  • Model options may vary based on your subscription plan
  • LiGo regularly updates available models as AI technology improves
  • Default settings are carefully selected based on extensive testing by our team
  • Some specialized models may be introduced for specific content types or industries

When to Adjust Model Settings

Consider reviewing and adjusting your model settings when:

  • You notice generation speed is slower than preferred
  • Generated content doesn't quite match your voice or expectations
  • You've updated your style packs significantly
  • You're working in a different content context (e.g., switching industries)
  • LiGo introduces new model options

Common Questions

Will changing models affect my saved style packs?

No, your style packs remain unchanged. Changing models only affects how the AI interprets and applies those style examples.

Which model is best for technical content?

For highly technical or specialized content, Claude often performs better at maintaining accuracy and proper terminology.

Can I use different models for different LinkedIn accounts?

Yes, if you switch between LinkedIn accounts in the extension, you can configure different model settings for each account.

Do some models work better with certain languages?

Yes, performance can vary by language. For non-English content, test different models to find the best performance for your specific language.

How often should I review these settings?

We recommend reviewing your settings quarterly or whenever LiGo introduces new model options.

Related Resources

Need more assistance?

If you couldn't find the answer you're looking for in this article, our support team is ready to help you.

Contact Support